“Ontology Without Trying” is best understood as a methodological proposal, not a recognized book, doctrine, or technical standard. It means asking ontological questions—what exists, what kinds of things matter, and how they relate—without first forcing experience or a working domain into a complete, permanent worldview. State a small set of assumptions, use them for the problem at hand, test their consequences, and revise them when they stop helping.
What ontology means
In philosophy, ontology is “the part of philosophy that studies what it means to exist,” according to the Cambridge English Dictionary (accessed September 27, 2026). Traditional ontological questions include:
- What kinds of things exist?
- Are properties, events, numbers, minds, or values real in the same sense as physical objects?
- What makes something the same thing over time?
- Which relationships are fundamental, and which are descriptions we impose?
These questions can lead to substantial metaphysical systems. The phrase without trying suggests a lighter commitment: investigate the assumptions you need, but do not pretend that a provisional map is the whole territory.
What “without trying” proposes
Because no identifiable author or work defines this exact title, the phrase should be treated as an interpretation. On that interpretation, it has four practical commitments.
Do these 3 things before closing this tab:
1Fix the driver behind crashes, sound loss and screen glitches2Repair Windows errors before they cause bigger problems3Scan for outdated or missing drivers - takes under a minute#1 Best Overall
Use assumptions openly
Every explanation begins somewhere. Instead of claiming to have discovered reality’s final categories, write down assumptions such as “people are agents,” “documents can refer to entities,” or “events have times.” Making them explicit lets others inspect and challenge them.
Prefer usefulness to premature completeness
A model can be good enough for a decision without settling every metaphysical dispute. A medical data model may need patients, diagnoses, tests, and dates; it does not need to resolve whether causality is ultimately fundamental.
Let evidence and consequences revise the model
If a category produces contradictions, excludes important cases, or makes ordinary work harder, change it. Provisional does not mean arbitrary: revisions should respond to evidence, clarified meanings, and the results of using the model.
Rank #2
Separate a working translation from a worldview
R. A. E. Olenius writes, “Think of what follows as a working translation, not a replacement theory,” and adds, “You are not being asked to adopt a worldview. You are being invited to notice a structure” (December 21, 2025). Those sentences capture the stance: notice recurring structure, name it carefully, and avoid turning the name into an unquestionable entity.
Outdated Drivers Are Slowing You Down
One free scan finds every outdated or missing driver and matches the right update for your exact hardware.Free scan · exact hardware matchPC Slower Than It Used to Be?
A free scan shows the junk files, broken settings and background clutter dragging Windows down - then fixes them in one click.Free scan · Windows 10 & 11Philosophical ontology and information-science ontology are different jobs
| Aspect | Philosophical ontology | Information-science ontology |
|---|---|---|
| Primary question | What does it mean to exist? | How should a domain’s concepts and relationships be represented? |
| Typical output | Arguments about beings, identity, reality, and fundamentality | A formal vocabulary of classes, properties, and restrictions |
| Success condition | Conceptual clarity and defensible argument | Consistent, shared, computable or interoperable domain descriptions |
| Revision | Driven by argument, objections, and explanatory power | Driven by domain knowledge, use cases, validation, and implementation feedback |
The Stanford University Protégé guide defines a technical ontology as “a formal explicit description of concepts in a domain of discourse,” including classes, properties, and restrictions on those properties. A technical ontology therefore does not automatically settle what exists in the universe; it specifies how a chosen domain is to be described.
Do you need an ontology before modeling a domain?
No. You need enough conceptual agreement to answer the immediate use case. Starting with a small, explicit model is often safer than designing an exhaustive hierarchy that nobody can maintain.
- State the use case. Identify the decisions, searches, integrations, or inferences the model must support.
- List recurring entities and events. Gather terms from domain experts, source systems, documents, and actual examples.
- Distinguish types from instances. “Invoice” may be a class; invoice 10482 is an instance. Record exceptions rather than silently forcing them into the wrong type.
- Define relationships and constraints. Specify properties such as issuedBy, expected value types, cardinality, and whether a relationship is optional.
- Test representative cases. Include ordinary records, edge cases, ambiguous terms, and contradictory data.
- Revise and document decisions. Keep a change history and explain why a term, boundary, or restriction changed.
Protégé’s guidance emphasizes that “there is no single correct ontology-design methodology.” Iteration is therefore a normal design feature, not evidence that the first version failed.
A practical provisional method
1. Keep an assumption ledger
For each important modeling choice, record the assumption, its evidence, its intended use, and what would falsify or weaken it. For example: “A subscription is active until its end date, unless a cancellation event is recorded.” This prevents an implementation rule from being mistaken for a universal truth.
2. Use the narrowest category that works
If “customer” is sufficient, do not immediately split it into legal person, account holder, payer, user, and beneficiary. Add distinctions when a requirement or observed conflict needs them.
Rank #4
3. Mark uncertainty instead of hiding it
Use labels such as unknown, disputed, inferred, or time-dependent where the source data cannot justify a stronger claim. A model that represents uncertainty is more reliable than one that converts gaps into false precision.
4. Check category boundaries
Ask whether two terms overlap, whether one is a subtype of the other, whether a relation has a clear direction, and whether the same word changes meaning between teams. Boundary tests expose many errors before they become data or code.
5. Review the model when the problem changes
A vocabulary designed for reporting may be inadequate for automation, legal audit, or cross-organization exchange. Revisit the assumptions when the purpose, data sources, or stakeholders change.
What this approach avoids
- False finality: treating a useful classification as the final inventory of reality.
- Unstated metaphysics: presenting design conventions as facts that require no argument.
- Category overgrowth: adding subclasses and properties before a real use case demands them.
- Semantic drift: allowing different teams to use one label for incompatible concepts.
- Tool-first modeling: letting a notation or software constraint decide the domain’s meaning.
How Putnam’s Ethics without Ontology relates
Hilary Putnam’s 2004 book is not the source or title of this phrase, but it offers a useful comparison. A Telos summary reproduces his argument that ethical objectivity need not depend on a special metaphysical realm: “I want to argue that the idea that ethical objectivity requires a special kind of metaphysical reality is a form of ‘ontological’ thinking that we can and should do without.”
Putnam’s point concerns ethical objectivity, whereas the present interpretation concerns how to work with ontological questions. The connection is methodological: both resist the assumption that a practice must first be grounded in an elaborate metaphysical inventory before its claims can be meaningful or assessable.
Questions to ask when using ontology provisionally
- What problem is this category solving?
- Which observations or requirements support it?
- What cases does it exclude or merge?
- Is this a claim about reality, a domain convention, or an implementation rule?
- What evidence would make us rename, split, merge, or remove it?
- Who needs to agree for the model to be useful?
Limits of the phrase
“Without trying” must not become an excuse to avoid definitions, evidence, or responsibility. A provisional model still needs clear terms, traceable decisions, tests, and owners. Nor does the phrase eliminate philosophical disagreement: choosing what to count as an entity, event, cause, or person can have ethical and political consequences. The advantage is transparency and revisability, not immunity from criticism.
Frequently Asked Questions
Can I think about reality without committing to a complete worldview?
Yes. You can state limited assumptions for a specific inquiry, examine their consequences, and revise them without claiming to have settled every metaphysical question.
Recommended Free Tools
What is ontology in AI and knowledge representation?
It is a formal description of a domain’s concepts, properties, and restrictions, used to give data and systems a shared vocabulary. It is narrower and more task-oriented than philosophical ontology.
Is “Ontology Without Trying” a known book or theory?
No identifiable canonical book, article, product, or attributed slogan establishes it as a named theory. It is most defensibly read as a provisional methodological phrase.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
